The team tested breath samples from more than 1,400 patients and identified 13 chemicals found in eight types of cancers, Crohn’s disease, Parkinson’s disease, pulmonary hypertension and other diseases. Each of those volatile organic compounds is present in varying amounts, forming a distinctive “fingerprint” for each ailment. “These odor signatures are what enables us to identify the diseases using the technology that we developed,” says research lead Prof. Hossam Haick.

To pick up the presence and ratio of the chemicals, the team built an “artificially intelligent nanoarray” called the Na-Nose. It uses specific sensors, like one made from gold nanoparticles and another that uses a network of carbon nanotubes, to sense the different compounds. The data is then analyzed by an artificial intelligence system, which takes into account age, gender and other factors, picking out the right affliction 86 percent of the time.